WUWT readers may remember this popular article from August 30th, 2010 New paper makes a hockey sticky wicket of Mann et al 98/99/08 and then The Team’s response RC’s response to McShane and Wyner: a case of orange cones which gave rise to Josh’s cartoon and this cartoon coffee mug:
Patrick Hadley writes in comments today:
OT – The McShane & Wyner discussion is now available at the Annals of Applied Statistics.
There is a lot of fascinating material to read there – the original paper, criticisms from the Hockey Team, support from others and, what seems to me at least, a brilliant rejoinder from McShane and Wyner.
I’ll say. Wow, this paper stirred up a statistical hornet’s nest, just have a look at the table of contents related to this paper. It reads like a who’s who of paleohockey. Each article is fully open, no paywalls; which I see as a testament to the journal integrity. If nothing else, read the editorial by Michael Stein which speaks to the entire table of contents.
There is so much here, I can’t even begin to dig into it all (It’s a workday for me) but if readers wish to place excerpts below of interest, I’ll do a follow up post with them. – Anthony
Discover more from Watts Up With That?
Subscribe to get the latest posts sent to your email.
Michael Stein blew it. He says “now is the time for individuals and governments to
act to limit the consequences of greenhouse gas emissions on the Earth’s
climate over the next century and well beyond” on the basis that the hypothesis “a large increase in the concentration of CO2 in the atmosphere has exactly zero effect on the global mean temperature” is implausible. As a mathematician, I have some sort of idea of exactly how large an amount is if it is not “exactly zero”. As a self-funded retiree and taxpayer in a ‘western’ country, I have some sort of idea how much these actions are going to cost me and those around me. As an inhabitant of various rural areas at various latitudes in my lifetime, I have some sort of idea of how beneficial warmth is.
However, there is a glimmer of hope in MS’s statement. He does not say “it is time to reduce greenhouse gas emissions, he says it is time to act to “limit the consequences of greenhouse gas emissions“. That is a whole new ball game, and one that no-one(?) except Bjorn Lomborg has addressed.
It seems that statisticians are the first professional group to have an open and transparent discussion of the numbers behind the science. What are the odds of that?
crosspatch – and why do you believe that? I lived just of the woods / nature for most of my lifetime, and I have never seen a cadaver of any larger size at all. The reason of course is that there are predators/scavengers taking advantage of the situation most of the time and/or it gets taken care of by smaller animals like bugs, insects, micro-organisms in no time. It does not stay there to supply nutrients for years to come.
So even IF an elk or whatever dies next to a tree, the number of samplings makes it statisticaly improbable to have any effect whatsoever.
Also, if you look up the shape and depths of the roots of pines in question here, EVEN IF a cadaver would remain in one spot for years, then the nutrient supply of a single tree would not be substantialy sourced from that cadaver simply because the portion of roots of that tree would be neglegible to the tree’s entire energy input.
All in all, that is something you just made up, has no substantiation at all and simply suits you for the moment. Elks would have to die under, and influence, a substantial amount of trees to invalidate tree ring reconstructions in the way you allege – but that does not match reality as we know it about animals dying and staying dead there in nature. So if we grant you an odd elk dying under a pine tree it had no influence – do you acknowledge that?
In Communist Eastern Europe, every scholarly article and book had to have a passage explaining how the results had already prefigured in the works of Friedrich Engels, or how they proved the validity of dialectical materialism. Authors usually appended this to the conclusion to keep it separate from the main body of the work, and it was therefore called the “red tail.”
These days another color is in vogue worldwide.
Plus ça change…
CP writes: Tree ring increases might also be CO2 proxies in certain species, particularly conifers.
No, they aren’t. It’s a simplistic theory that fails to consider the complexity of tree diameter growth. The confounding complexity is well-respected, and has led to a search for special individual trees in locations on the extreme edge of tree survival. That’s why in a world with trillions of trees, only a handful have been “qualified” as proxies for temperature and/or CO2-related growth anomalies. And why the data from those rare, special trees is totally questionable as a proxy for anything.
How come editors and journalists don’t inadvertently acquire a little knowledge or perhaps wisdom enough to realise their ignorance. Fight the greenhouse gases by all means but remember they are water vapour 93%, methane 3%, carbon dioxide 3% and a few odds and ends. The reason variations of CO2 are of no consequence are that the absorption spectra for infrared are fully saturated. 65 million years ago the CO2 level was 3000 ppm which was followed by a progressive fall in temperature for 20 million years until Drake’s Passage opened enough for the circum-polar current to become established, freeze Antarctica and establish the Ice Ages cycle mainly in sync with one of the Milancovich Cycles. So what was the other question? Did I forget to mention sooty chimneys and SUVs? Geoff Broadbent
boballab says:
December 13, 2010 at 11:01 am
M&W’s rejoinder to SMR would fall into the category that Phil Jones calls “going to town on”.
Compounding matters, SMR implement their allegedly objective criteria in non-standard and arbitrary ways and several times in error4. When correctly implemented, the number of principal components retained varies across each ”objective” criterion from two to fifty-seven. Using ten principal components, therefore, can hardly be said to induce ”statistical overfitting” claimed by SMR.
Ouch! That will leave a mark.
======================================================
Lol, it wasn’t just there that they ‘laid the smack down’.
We fault many of our predecessors for assiduously collecting and presenting
all the facts that confirm their theories while failing to seek facts that contradict
them. For science to work properly, it is vital to stress one’s model
to its fullest capacity (Feynman, 1974).
That was well-stated and long overdue in a science journal.
We are able to show, by brute force computation, that our results are invariant
to these choices. Furthermore, as stated in our paper, we implemented
many of these proposals prior to submission (for discussion of variations
originally considered and justification of our choices, see Section 3.7 for
the Lasso; footnote 8 for thirty year blocks; Section 3.4 for interpolation;
and Section 3.6 for calibration to local temperatures). In contrast, we credit
McIntyre and McKitrick (MM) for pointing out the robustness of these results
and Kaplan for actually demonstrating it by using Ridge regression
in place of the Lasso (see Kaplan Figures 1 and 2). We direct the reader to
our SI where we perform the same tests (1) for a plethora of methods (including
the elastic net called for by HU and the Noncentral Lasso called for
by Tingley) (2) using thirty and sixty year holdout blocks (3) using both interpolated
and extrapolated blocks and (4) fitting to the local temperature
grid as well as CRU when feasible. Once again, the results demonstrated
by Figures 9 and 10 of our paper are robust to all of these variations.
One has to read the context of this statement, but, it shows either the lack of reading comprehension of their critics or a total lack of candor. I had already known they’d used other methods. How? I read their paper. I guess the critics figured that wouldn’t be able to back it up in a science journals. I think its called ‘projection’.
Second, we take the data as given and do not account for uncertainties, errors, and biases in selection, processing, in-filling, and smoothing of the data as well as the possibility that the data has been ”snooped” (subconsciously or otherwise) based on key features of the first and last block.
Thus, we take Rougier’s characterization
of our model (”perfectly reasonable ad-hockery”) as high praise.
Their rejoinder is a work of art!
……back to reading……
In ‘The rejoinder’ (last link of the article above) on Page 16- first complete para- and the ‘Conclusions,’ put a bonfire of the vanities under Manns data
tonyb
Rougier has introduced a new term “ad-hockery”
(imsart-aoas ver. 2010/09/07 file: AoASMcShane_Rougier.tex date: September 9, 2010)
Quote from J.C. ROUGIER – “I am bemused by section 5. First, let us be very clear that this is is not a “fully Bayesian” analysis. What we have here is a normalised likelihood function over and masquerading as a posterior distribution, in order to implement a sampling procedure over the model parameters. This seems a perfectly reasonable ad-hockery (although a Normal Inverse Gamma conjugate analysis would be more conventional, see O’Hagan and Forster, 2004, ch. 11), but to call it “fully Bayesian” is stretching the point. ” End Quote.
I love it – a new term for Mann’s efforts – “ad-hockery” – LOL
Gary Palmgren says:
December 13, 2010 at 9:59 am
The use of the precautionary principle with regards to CO2 emissions causing damaging overheating has no basis in the geological history. It is cold we dread. The misery during the little ice age is well documented. Where does this causal acceptance that warming is bad come from?
###
’tis not being dead that my mortal dread,
its an icy grave that pains,
so I’m asking you, to promise true,
too cremate my last remains.
The warmth == BAD narrative is needed to scare people into giving up their freedom to become slaves to the state. A good economy makes for a hard to control populace. If the masses ever found out that 3 or 4 degree C increase in temperature is a “good thing”, all hell would break loose. Its funny, that we call those times of warm climate, “Golden Ages”.
There are lots of choice quotes in the Rejoinder:
Ouch!
Garey, Lattitude:
Where does this causal acceptance that warming is bad come from?
I think you mean ‘casual.’ It’s a good rhetorical question, pointing to an obvious answer: ‘from lack of due diligence.’
The interesting thing, to me, were the negative reviews by statistitians. Most ( not all) found MW lacking in statistical skill. By inference they also criticised McIntyre.
I know its hard, but the lukewarmers and warmists need to grow some cahoonas – accept that they are starting from a flawed ‘view’ and press the Big Reset Button.
In the same way as the weird alarmists advocate the precautionary principle – the sceptics should advocate a precautionary reset!
With a good degree of proper co-ordination, co-operation and a return to the principles of the scientific method – the warmists could ‘keep’ some credibility. Everyday, as they continue to shout out their biased conclusions, they fall foul of all that is scientific.
Any warmists or lukewarmers working in the field should seriously consider extending the hand of scientific friendship and ‘peace’ and start over. Looking at it another way, assume they are right and AGW is damning us all to hell, by resetting and reworking they have the chance to prove it to all the sceptics properly and gain many more converts! (pretty much all of us, eh?)
Its like having a maths teacher who needs to keep explaining the same obscure point over and over again but without working it through with a student, the student just cannot get it. Moreover, the teacher cannot understand why the student doesnt ‘get it’ but a proper work through would help BOTH improve their understanding. Ok – this is simplistic, but it is entirely logical. If Mann et al, really believe their data, they will ‘work through’ it in great detail with the likes of Mcshane, etc- and prove it is correct. Why don’t they? Why is it that it always seem to be the sceptics ‘challenging’ the concensus?
The precautionary principle can work both ways if explained correctly?
lol, is Gavin really that thick?
From SMR—-“Using their reconstruction, MW nonetheless still found recent warmth to be unusual in a long-term context: they estimate an 80% probability that the decade 1997-2006 is warmer than any other for at least the 50 past 1000 years.”
lol, when MW was released, I went to RC and told him this shouldn’t be interpreted as an assertion by MW.(one of the very few times one of my posts were allowed there.) Of course, he argued with me.
In the rejoinder, M&W make it explicitly clear…..“It is thus highly
improbable that they would be able to detect such high levels and sharp
run-ups if they indeed occurred in the more distant past. That is, we lack
statistical evidence that the recently observed rapid rise in temperature is
historically anomalous.”—— which is almost exactly what they stated in the original paper.
Great news from Britain! TV comedy shows are beginning to get a laugh out of Global Warming. On the BBC, Armstrong and Miller did a sketch showing a man looking out at lousy weather and moaning, “Huh! Whatever happened to Global Warming, eh”, followed by threats of imprisonment, and government leaflets explaining the difference between weather and climate.
Let’s hope that the tide of ridicule will rise so high that belief in Global Warming will be something to confess to. I wonder if this is happening elsewhere in the English-speaking world.
John McManus says:
December 13, 2010 at 1:59 pm
The interesting thing, to me, were the negative reviews by statistitians. Most ( not all) found MW lacking in statistical skill. By inference they also criticised McIntyre.
========================================================
Could you provide the links to the discussion papers you’re reading?
Happy people in paleoclimate
Now there is a break from the usual bleary tree-ring photo we are used to staring at all the freakin’ time. I mean this one.
🙂
From the Rejoinder-
If that isn’t calling the into question whether we can ‘extract’ a temperature signal out of a proxy, I don’t know what is. Wow!
And in the same sentence, they take a good swipe at the methods employed by the climate community-
The response from the Team will be most interesting…..
Hello? Stop! stop! stop ! Have many of you gone paranoid?
Cant see the difference betweeen scientists and advocates no more? Friends and enimies?
This is the best initiative so far to really take an audit of the roots of the scientific problems in climate science we all sceptics beeen reacting to.If the editor hadnt ended his introduktion with the words on global warming his inroduction would be considered to be an declaration of war against Climate science.
Its obvious that they now have realized that the hole climate science backbone is made out of statistic flaws and unouthorised freedom to interpretation and speculation far away from established statistical skills.They have the skills and methology to calculate on how big is the room fore pure speculation is.IUts the matematic constaples of order and diciplin that has finnaly tuned up.See them as the adults entering the climateresearch caotic playground.
This work and intrest from these guys are extremely valuble and when you read the answer from McShane and Wyner the enomous gap in level of statistic skills and understanding between these guys and the “team” youll find the gap embarrasin to the team and it becomes obvious whos the “daddys” are.
Before embarking on our discussion of their work, we must mention that,
of the five discussants who performed analyses (DL, Kaplan, SMR, Smer-
don, and Tingley), SMR was the only one who provided an incomplete and
generally unusable repository of data and code. The repository created by
SMR specifically for this discussion was, like that of the other four discus-
sants, graciously provided and quite usable. However, we lacked clear and
easily implementable code (i) to fit RegEM EIV ourselves and (ii) to draw
new temperatures and pseudoproxies from their simulation model. Code
for these purposes is archived by Mann at:
http://www.meteo.psu.edu/
∼mann/PseudoproxyJGR06/
http://www.meteo.psu.edu/
∼mann/supplements/MultiproxyMeans07/
Among other things, the RegEM EIV fitting procedure cannot be executed
by a straightfoward function call as is typical for statistical code libraries.
Rather, the archives consist of a large number of files layered on top of one
another and, despite a major effort on our part, we were unable to replicate
published results within the publication time constraints of this rejoinder.
More errors in Schmidt,Mann and Rutherford
Fortunately, we are able to use the data and code provided to us to re-
but SMR’s findings. Before proceeding, however, we must note a troubling
problem with SMR Figure 2. Visual inspection of the plots reveals an errant
feature: OLS methods appear to have non-zero average residual in-sample!
Upon examining the code SMR did provide, we confirmed that this is in-
deed the case and discovered the models were fit incorrectly. The culprit,
ironically, is an improper centering of the fitted values.
I’m not a statistician but reading the rejoinder was awesome. I love how they dismantled SMR piece by piece!
+1 for me… the editorial reads with a distinct flow of integrity until the last line where Stein abandons all reason. Why put so much emphasis on the uncertainties and then plunge willy-nilly into the precautionary approach regardless? Basically he has sidelined the entire purpose of the editorial of saying that more statistical input is required in the science, by saying it is all irrelevant for policy-making anyway… just go ahead and pretend CAGW is true for all policy purposes. Pathetic.
Ian W
“. . . all the metrics appear to be on ‘temperature’ which is NOT the correct metric for ‘atmospheric heat content’ which is what they all claim to be measuring. It really doesn’t matter how elegant and accurate the statistics are, if they are the statistics of the incorrect metric.”
Absolutely spot on – the point is crucially important to the whole exercise. How any serious scientists got duped by people who use and misuse high school statistics as the hockey team and the rest of them have is beyond me.